ORIGINAL RESEARCH Treating to Meet NCEP-Recommended LDL Cholesterol Concentrations with Atorvastatin, Fluvastatin, Lovastatin, or Simvastatin in Patients with Risk Factors for Coronary Heart Disease

Donald Hunninghake, MD; Rebecca G. Bakker-Arkema, MS; James P. Wigand, MD; Margaret Drehobl, MD; Helmut Schrott, MD; James L. Early, MD; Paul Abdallah, MD; Scott McBride, MA; and Donald M. Black, MD Minneapolis, Minnesota; Ann Arbor, Michigan; Richmond, Virginia; San Diego, California; Iowa City, Iowa; Wichita, Kansas; and Seattle, Washington

BACKGROUND. Our study compared use of atorvastatin, fluvastatin, lovastatin, and simvastatin for lowering low-density lipoprotein (LDL) cholesterol concentration in patients at risk for coronary heart disease (CHD). The goal was to reach the LDL cholesterol levels recommended by the National Cholesterol Education Program (NCEP).

METHODS. A combined total of 344 men and women took part in this 54-week, multicenter, open-label, random­ ized, parallel-group, active-controlled, treat-to-target study. Patients were selected on the basis of their LDL cho­ lesterol concentration and their risk for CHD. During treatment, doses were titrated at 12-week intervals to a maximum of 80 mg per day of atorvastatin and lovastatin, or 40 mg per day of fluvastatin and simvastatin, with colestipol added if necessary to attain the NCEP-recommended LDL cholesterol concentration.

RESULTS. At the starting dose, atorvastatin decreased plasma LDL cholesterol significantly (P <.05) compared with the other reductase inhibitors, and the percentage of patients reaching target LDL cholesterol concentration at the starting dose was significantly greater in the atorvastatin group (P <.05). Overall, a significantly (P <.05) greater percentage (95%) of atorvastatin-treated patients achieved target LDL cholesterol concentration. The safety profile was similar among all reductase inhibitors tested.

CONCLUSIONS. At the starting dose, a significantly (P <.05) greater percentage of atorvastatin-treated patients at risk for CHD reached the target LDL cholesterol concentration than patients treated with other reductase inhibitors.

KEY WORDS. Cholesterol; atherosclerosis; coronary heart disease. (J Fam Pract 1998; 47:349-356)

everal clinical trials have demonstrated that approach to treating high blood cholesterol in adults.3 elevated serum low-density lipoprotein (LDL) These NCEP guidelines are based on the patient’s exist­ cholesterol is associated with increased risk of ing LDL cholesterol concentration and risk for CHD. coronary heart disease (CHD), and that lower­ The expert panel recommends lipid-lowering drug ing serum LDL cholesterol levels reduces the treatment if, after an attempt at dietary intervention, Slikelihood of new coronary events and associated mor­LDL cholesterol remains >190 mg/dL in patients with tality.12 In 1993, the Adult Treatment Panel (ATP-II) less than 2 CHD risk factors or >160 mg/dL in patients report of the National Cholesterol Education Program with 2 or more CHD risk factors. (NCEP) outlined an updated systematic clinical Atorvastatin is a new 3-hydroxy-3-methylglutaryl- coenzyme A reductase inhibitor that has been shown to Submitted, revised, July 15, 1998. lower LDL cholesterol 41% to 60% over its effective From the University of Minnesota (D.H.); Wigand, dose range.4"6 In our study, we prospectively evaluated Wasserman, Burris & Young Medical Associates (J.P. W.); the ability of atorvastatin to treat patients at risk for Center for Health Care Medical Associates (M.D.); CHD to their NCEP-recommended target LDL choles­ University of Iowa (H.S.); University of Kansas School of terol concentration. Atorvastatin was compared with Medicine (J.L.E.); Minor & James Medical (P.A.); and Parke-Davis Pharmaceutical Research, Division of Warner- three frequently prescribed reductase inhibitors: fluvas­ Lambert Co (R.G.B., S.M., D.M.B.). Requests for reprints tatin, lovastatin, and simvastatin. Our study is the first should be addressed to Rebecca G. Bakker-Arkema, Parke- direct comparison between these agents in a dyslipi- Davis Pharmaceutical Research, 2800 Plymouth Rd, Ann demic population being treated to NCEP-recommended Arbor, MI 48105-1047. LDL cholesterol concentrations.

1998 Appleton & Lange/ISSN 0094-3509 The Journal of Family Practice, Vol. 47, No. 5 (Nov), 1998 3 49 TREATING TO MEET NCEP-RECOMMENDED LDL CHOLESTEROL CONCENTRATIONS

METHODS grouped according to their number of risk factors as described by the NCEP: less than 2 risk factors for CHD Study Design and mean cholesterol >190 mg/dL or 2 or more CHD risk This 54-week, open-label, randomized, parallel-group, factors and a mean LDL cholesterol value of >160mg/dL active-controlled, treat-to-target study in patients at risk Mean LDL was calculated using the Friedewald formula for cardiovascular disease consisted of three phases: an at weeks -4 and -2.8 A total of 344 patients were random­ optional 8-week screening and dietary assessment ized to the study: 82 with fewer than 2 risk factors and phase, a 4-week lead-in phase, and a 54-week open-label 262 with 2 or more risk factors. The distribution of treatment phase. The study ran from March 1995 to patients within risk categories was similar across treat­ October 1996. We recruited patients from 24 facilities ment groups. Patient exclusions are listed in Table 1. reflecting a broad range of medical practices in the Eligible patients were randomized to 1 of 4 treatment United States, including 12 primary (eg, internal medi­ groups using starting doses of atorvastatin 10 mg per cine, family practice) and 12 secondary (eg, cardiology, day, fluvastatin 20 mg per day, lovastatin 20 mg per day, lipid) centers. Identical protocols were reviewed and or simvastatin 10 mg per day. Lipids were measured at 6- approved by an institutional review board. Patients gave week intervals, and dose titration occurred at 12-week written informed consent before participating. Women intervals (weeks 12, 24, 36 and 48), until the patient of nonchildbearing potential, and men and women at reached the LDL cholesterol concentration recommend­ risk for CHD, aged 18 to 80 years, and with a body mass ed by the NCEP ATP-II. index <32 kg/m2 were screened for eligibility with a The dose of reductase inhibitor could be increased to physical examination, a medical history, a clinical labo­ a maximum of 80 mg per day for atorvastatin, 40 mg per ratory evaluation, a urinalysis, and a lipid profile. day for fluvastatin, 80 mg per day for lovastatin, and 40 Eligible patients were required to adhere to the NCEP mg per day for simvastatin (the maximum approved Step I or Step II diet or a similar diet during the study.3 7 dose at the time of the study). If target LDL cholesterol We allowed the lipid levels of patients beginning dietary concentration was not achieved at the maximum dose of modification to stabilize over an 8-week period before the reductase inhibitor, colestipol was added to the the patient entered the 4-week lead-in phase (see figure patient’s regimen, initially at 5 g twice daily. On the basis 1). The lead-in phase was used to further evaluate the of the patient’s response, the colestipol dose could be patient’s eligibility and to establish baseline values for increased by 5 g twice daily after a 12-week interval, to a study parameters. Fasting triglyceride values of <400 maximum of 20 g per day. If the patient could not toler­ mg/dL were necessary for inclusion. Patients were also ate the drug, the dose could be decreased or totally withheld. - TABLE 1 ______Once the target LDL cholesterol concentration was Exclusion Criteria for Study Comparing Atorvastatin, reached, the patient’s dose of mono- or combination- Fluvastatin, Lovastatin, and Simvastatin therapy was maintained with no further titrations regardless of subsequent LDL cholesterol values. Lipids •Hypersensitivities to reductase inhibitors or bile acid seques­ were measured at 12-week intervals until the end of the tering resins study. At weeks 0 and 54, patients completed a 24-hour •Prohibited medications, such as lipid regulating drugs not dietary diary. An investigator reviewed the contents of prescribed in the protocol, immunosuppressive agents, and the diaries for completeness and sent them to the drugs known to be associated with rhabdomyolysis in combi­ Chicago Center for Clinical Research, where a food nation with reductase inhibitors (ie, cyclosporine, ery­ record rating score was calculated.” Noncompliant thromycin) patients (those who took less than 80% of their pre­ •Pregnancy or breast-feeding scribed medication) were counseled but were not •Secondary causes of hyperlipoproteinemia, such as uncon­ dropped from the study. trolled hypothyroidism, nephrotic syndrome, severe renal dys­ function, or uncontrolled type 1 or type 2 diabetes mellitus Efficacy and Safety Measurements Clinical and safety evaluations and lipid profiles were •Active liver disease or hepatic dysfunction managed through an accredited, standardized, central •Myocardial infarction, coronary angioplasty, coronary artery laboratory (Pacific Biometrics Research Foundation, bypass graft, or severe or unstable angina pectoris within 1 Seattle, Washington). Serum samples for lipid profiles month of screening were collected after a minimum 12-hour fast, and blood •Participation in another clinical study in which study medica­ samples for lipid profiles were drawn between 6 and 18 tion was received within 30 days of screening for this study hours postdose. •Significant abnormalities that the investigator judged could We evaluated total cholesterol, LDL cholesterol, compromise the patient’s safety or successful participation in triglyceride, and high-density lipoprotein (HDL) choles­ the study terol levels. Triglyceride levels were determined enzy­ matically with the Hitachi 737 analyzer.1" Plasma HDL

3 5 0 The Journal of Family Practice, Vol. 47, No. 5 (Nov), 1998 TREATING TO MEET NCEP-RECOMMENDED LDL CHOLESTEROL CONCENTRATIONS

cholesterol concentration was determined enzymatically group would detect a difference of 1.5 visits with 80% after LDL and very-low-density lipoprotein (VLDL) cho­ power. Estimating a 15% dropout rate yielded a lesterol were selectively removed from the plasma sam­ requirement of approximately 80 patients per group. ple by heparin and magnesium chloride precipitation.'1 For triglyceride levels <400 mg/dL,8 LDL cholesterol con­ Efficacy Descriptive statistics were prepared for all centration was estimated using the Friedewald formula. baseline demographic and lipid variables. Statistical For triglycerides >400 mg/dL, LDL cholesterol levels tests were performed on data from weeks 12 (starting were determined by ultracentrifugation.12 dose), 24 (before colestipol adjuvant therapy), and 54 A full clinical laboratory evaluation was performed at (end of study). Efficacy analyses were performed on screening, at randomization, and at the end of the study, data from an intent-to-treat population, defined as all while evaluations for safety (alanine aminotransferase patients who received at least one dose of study med­ [ALT], aspartate aminotransferase [AST], and creatine ication and who had at least one lipid measurement phosphokinase [CPK]) were done at all intervening vis­ taken after randomization. All statistical tests were its. Adverse events were recorded at each clinic visit and two-sided and conducted at the 5% level of signifi­ up to 15 days after treatment ended. Associated adverse cance. events were those the investigator judged as definitely, Analysis of covariance was performed to compare probably, or possibly related to treatment, as well as the effects of the 4 treatments on the percent change those that had an unknown relationship to treatment or from baseline in LDL cholesterol levels, total choles­ insufficient information available for evaluation. terol, triglycerides, and HDL cholesterol, at weeks 12, 24, and 54. The model included the effects of treatment, Statistics center, CHD risk factors, and the baseline lipid value as Power The sample size calculation was based on a a covariate. A Dunnett test was used to perform pair­ two-sided t test at the 5% level of significance. Patients wise comparisons between atorvastatin and the other were stratified by CHD risk factors so that patients treatments. The last available postrandomization lipid with 2 or more CHD risk factors were 75% of the total measurement was carried forward to explain missing sample and patients with less than 2 CHD risk factors observations. were 25% of the total sample. The percentage of For determining the percentage of patients reaching patients reaching target at each visit was estimated. target LDL cholesterol concentration, responders were The estimate of the mean number of visits needed to those patients with less than 2 CHD risk factors whose reach target in each treatment group yielded a differ­ mean LDL cholesterol was <160 mg/dL or patients with ence of 1.5 between atorvastatin and simvastatin. An 2 or more CHD risk factors whose mean LDL cholesterol estimate of 3.0 for the standard deviation was used for was <130 mg/dL. Patients were counted as responders at determining sample size. A sample of 68 patients per weeks 12, 24, and 54 if they had achieved that status at

FIGURE 1

Study Design

mg atorvastatin QD QD QD ------; ------1 QD 10 g colestipol (5 g BID)

10 mg simvastatin. 20 mg simvastatin 40 mg simvastatin 40 mg simvastatin QD 40 mg simvastatin QD j QD QD QD + + Diet 10 g colestipol (5 g BID) 20 g colestipol (10 g BID)

1 20 mg lovastatin 40 mg lovastatin 80 mg lovastatin* 80 mg lovastatin* . 80 mg lovastatin* ------QD------h QD + + 10 g colestipol (5 g BID) 20 g colestipol (10 g BID)

20 mg fluvastatin 40 mg fluvastatin I 40 mg fluvastatin QD 40 mg fluvastatin QD * 20 g colestipol (10 g BID) QD 10 g colestipol (5 g BID)

8-Week (optional) 4-Week Dietary 4-Weeks 54-Week Treatment Phase Dietary Assesment Phase Lead-In Phase t t Off prior 1 Screen t dyslipidemic Randomization 6 weeks 12 weeks 18 weeks 24 weeks 30 weeks 36 weeks 42 weeks 48 weeks 54 weeks medication

Possible titration points T T T t

*40 mg BID.

The Journal of Family Practice, Vol. 47, No. 5 (Nov), 1998 3 5 1 TREATING TO MEET NCEP-RECOMMENDED LDL CHOLESTEROL CONCENTRATIONS

TABLE 2

Baseline Patient Characteristics

Atorvastatin Fluvastatin Lovastatin Simvastatin All Patients Characteristic (n = 86) (n = 85) (n = 86) (n = 87) (N = 344)

Sex, no. (%) Men 40 (47) 34 (40) 46 (53) 41 (47) 161 (47) Women 46 (53) 51 (60) 40 (47) 46 (53) 183 (53)

Race, no. (%) White 80 (93) 75 (88) 76 (88) 81 (93) 312 (91) Other 6(7) 10 (12) 10 (12) 6(7) 32 (9)

Age, years Median 55 55 55 58 56 Range 20 to 80 28 to 77 27 to 78 27 to 78 20 to 80 Mean (SE) 55 (1.4) 56 (1.3) 55 (1.2) 57 (1.3) 56 (0.6)

CHD risk factors, no. (%) <2 22 (26) 21 (25) 20 (23) 19 (22) 82 (24) >2 64 (74) 64 (75) 66 (77) 68 (78) 262 (76)

Lipid values (mg/dL), mean (SE) LDL-C 205 (4.3) 201 (4.1) 206 (3.9) 210 (5.5) 205 (2.3) Total cholesterol 286 (4.6) 286 (3.9) 290 (4.4) 292 (5.8) 289 (2,4) Total TG 190 (7.4) 209 (8.8) 205 (8.2) 201 (8.4) 201 (4.1) HDL-C 42 (1.2) 43 (1.1) 43 (1.2) 42 (1.0) 43 (0.6)

CHD denotes coronary heart disease; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; TG, triglycerides.

the evaluation visit or at any previous titration visit, or weakness were considered important laboratory devi­ regardless of whether that status had been maintained ations because of the increased incidence of these labo­ throughout the study. All patients who withdrew without ratory events with reductase inhibitors. having met responder status were counted as non­ responders. RESULTS We used the Cochran-Mantel-Haenszel analysis stratified by center and CHD risk factors to compare Demographics the percentage of patients in treatment groups achiev­ A total of 344 patients were randomized into the study; ing target LDL cholesterol concentration at each eval­ 82 with less than 2 risk factors and 262 with 2 or more uation point. The amount of the reductase inhibitor risk factors. Twelve primary care centers enrolled 194 and the amount of colestipol required by each treat­ patients (56%); 6 lipid centers, 90 patients (26%); and 6 ment group to reach target LDL cholesterol concentra­ cardiology centers, 60 patients (17%). The distribution of tion by week 54 were compared using analysis of vari­ patients within risk categories was similar across the ance with a model that included the effect of treat­ treatment groups. There were no apparent differences in ment. Between treatments, the generalized Wilcoxon baseline characteristics across the groups (see Table 2). test without stratification by center or CHD risk fac­ Patients were primarily white (91%), with a mean age of tors was used to compare the time to reach target LDL 56 years (range: 20 to 80). Mean baseline lipid values cholesterol concentration by week 54. were similar across the groups.

Safety Compliance All patients who received study medication were evalu­ Compliance, based on tablet counts at each visit, was ated for safety. Adverse events and laboratory deviations high for all treatment groups. It ranged from 86% to 99% outside the normal range were recorded at clinic visits. for the atorvastatin group, 75% to 99% for the fluvastatin Adverse events were summarized to assess the adverse group, 71% to 96% for the lovastatin group, and 71% to event rate for the treatment groups. An increase in 96% for the simvastatin group. Food record rating scores transaminase levels (ALT or AST) greater than 3 times across treatment groups were similar at baseline and the the upper limit of normal (ULN) and an increase in CPK end of study, and values were similar within each treat­ greater than 10 times ULN with muscle pain, tenderness, ment group at baseline and at the end of study.

3 5 2 The Journal of Family Practice, Vol. 47, No. 5 (Nov), 1998 TREATING TO MEET NCEP-RECOMMENDED LDL CHOLESTEROL CONCENTRATIONS

Disposition and Exposure At week 54, atorvastatin significantly lowered LDL Ninety percent of patients in the atorvastatin group com­ cholesterol levels from baseline to a greater extent than pleted the study compared with 88% for fluvastatin, 83% fluvastatin and lovastatin, as monotherapy or in combi­ for lovastatin, and 89% for simvastatin. At the end of the nation with colestipol. Atorvastatin-treated patients had study, 70% of atorvastatin patients were exposed to a significantly greater reductions in total cholesterol lev­ maximum dose of 10 mg, and only 2% were exposed to els than fluvastatin-, lovastatin-, or simvastatin-treated combination therapy. In contrast, 20% of fluvastatin patients. HDL cholesterol levels increased similarly patients were exposed to a maximum dose of 20 mg, and across all treatment groups. 67% were exposed to combination therapy; 40% of lovastatin patients were exposed to a maximum dose of NCEP Target LDL Cholesterol 20 mg, and 24% were exposed to combination therapy; Concentration and 43% of simvastatin patients were exposed to a max­ After 12 weeks of treatment, when all patients were imum dose of 10 mg, and 24% were exposed to combi­ given the starting dose of treatment, a significantly nation therapy. (P <.05) greater percentage of atorvastatin-treated patients (71%) achieved the target LDL cholesterol con­ Efficacy centration compared with 16%, 34%, and 41% for fluvas­ Efficacy data are summarized in Table 3. Statistical eval­ tatin-, lovastatin-, and simvastatin-treated patients, uation of week 12 data and week 24 data indicated that respectively. Results were similar at week 24. By the end atorvastatin decreased LDL cholesterol and total choles­ of the study the percentage of patients reaching the tar­ terol levels to a significantly greater extent than fluvas­ get LDL cholesterol concentration remained significant­ tatin, lovastatin, or simvastatin. HDL cholesterol con­ ly (P <.05) higher for atorvastatin-treated (95%), than centration increased to a similar extent across all treat­ fluvastatin-treated (60%), lovastatin-treated (77%), or ment groups. simvastatin-treated (83%) patients. Cumulative response

_ TABLE 3

Mean Percent Change* from Baseline in Lipid Parameters

Atorvastatin Fluvastatin Lovastatin Simvastatin Lipid Parameter (n = 85) (n = 82) (n = 83) (n = 87)

LDL cholesterol Week 12, % change (SE)t -36 (1.3) -14|| (1.3) -23|| (1.3) -29|| (1.3) Week 24, % change (SE)t -35 (1.3) -1911 (1-4) -24|| (1.4) -29|| (1.3) Week 54, % change (SE)§ -36 (1.6) -2211 (1.6) -28|| (1.6) -33 (1.6)

Total cholesterol Week 12, % change (SE)f -28 (1.0) -1111 (1.0) -17|| (1.0) -21II (1.0) Week 24, % change (SE)f -27 (1.0) -14|| (1.0) -18|| (1.0) -211| (1.0) Week 54, % change (SE)§ -28 (1.2) -15|| (1.2) -211| (1.2) -24|| (1.2)

Triglycerides Week 12, % change (SE)f -16(3.1) -7(3.1) -5|| (3.1) -5|| (3.1) Week 24, % change (SE)f -18 (2.7) -9|| (2.7) -13 (2.7) -15(2.7) Week 54, % change (SE)§ -20 (3.4) 2|| (3.4) -16 (3.4) -11 (3.4)

HDL cholesterol Week 12, % change (SE)f 4(1.3) 5(1.3) 5(1.3) 6(1.3) Week 24, % change (SE)t 5(1.4) 8(1.4) 9(1.4) 9(1.4) Week 54, % change (SE)§ 9(1.4) 6(1.5) 10 (1.5) 11 (1.4)

HDL denotes high-density lipoprotein; LDL, low-density lipoprotein; SE, standard error. 'Least squares mean provided for percent change based on analysis of covariance (ANCOVA) model with effects due to treatment, center, coronary heart disease risk factors, and the baseline lipid value as a covariate. tThese data represent the use of drugs at the starting dose: atorvastatin 10 mg, fluvastatin 20 mg, lovastatin 20 mg, and simvastatin 10 mg. (These data represent the use of drugs across dosage ranges: atorvastatin 10-20 mg, fluvastatin 20-40 mg, lovastatin 20-40 mg, and simvastatin 10-20 mg. §These data represent the use of drugs across the full dosage range and in combination with colestipol adjuvant therapy. ilSignificantly different from atorvastatin (P <.05); Dunnett test.

The Journal of Family Practice, Vol. 47, No. 5 (Nov), 1998 3 5 3 TREATING TO MEET NCEP-RECOMMENDED LDL CHOLESTEROL CONCENTRATIONS

for the fluvastatin group (34%) and the lovastatin group Cumulative percentage of patients reaching the (24%). The simvastatin group experienced slightly more NCEP-recommended LDL cholesterol concentration.* related adverse events (18%) than the atorvastatin group. Digestive system events were the most frequently reported treatment-associated adverse events, occurring with a higher incidence among fluvastatin- and lovas- tatin-treated patients than those receiving the other reductase inhibitors. Treatment-associated adverse events led to the withdrawal of 3% of the patients treat­ ed with atorvastatin, 4% treated with fluvastatin, 8% treated with lovastatin, and 5% treated with simvastatin. One patient experienced a serious adverse event (acute pancreatitis) considered associated with treatment while taking fluvastatin. Two patients randomized to receive lovastatin treatment died during the study. The study investigator did not associate either of these deaths with the study drug. Some minor sporadic elevations in ALT and AST were noted in all treatment groups. Changes in remaining parameters were not clinically meaningful and showed *Note: The median dose for each treatment group at week 54 (with no treatment-associated trends. One lovastatin-treated the last available dose carried forward) was atorvastatin 10 mg per patient experienced ALT values over 3 times ULN. The day, fluvastatin 40 mg per day, lovastatin 40 mg per day, and study medication was temporarily stopped and the ALT simvastatin 20 mg per day. level returned to within normal range. over time is shown in Figure 2. DISCUSSION Patients in the atorvastatin treatment group required significantly (P <.05) less time to reach the target LDL Three hundred forty-four patients were enrolled in 12 cholesterol concentration than patients in the other primary and 12 secondary care facilities participating reductase inhibitor treatment groups. Atorvastatin-treat- in our study. The majority of patients were enrolled by ed patients had a median time to response of 85 days the primary care and lipid centers, supporting the fact compared with 269, 232, and 173 days for fluvastatin, that many patients at risk for developing CHD are lovastatin, and simvastatin, respectively.13 Overall, target being treated at these facilities. In addition, although LDL cholesterol concentration was achieved in the ator­ the cardiology centers were designated as secondary vastatin treatment group with significantly (P <.05) less care facilities, their patient recruitment (approximate­ reductase inhibitor and with significantly (P <.05) less ly 50% of the other sites) reflected a reasonable pri­ colestipol combination therapy than required in the mary care patient base. other treatment groups. Study patients taking atorvastatin achieved their The percentage of patients experiencing adverse NCEP-recommended target LDL cholesterol concen­ events was similar for all treatments. Related adverse tration more often than patients using the other reduc­ events were summarized at weeks 24 and 54. Week 24 tase inhibitors, with fewer office visits and less combi­ was the last visit at which all patients were still taking nation therapy. Consequently, mean total cost of care monotherapy. Thus, adverse event summaries through to reach the target concentration was lowest with 24 weeks were used to compare profiles between treat­ atorvastatin.13 ment groups without confounding effects from The percentage of related adverse events reported by colestipol. Daily dose ranges at week 24 were 10 to 20 at least 2% remained relatively consistent from week 24 mg for atorvastatin, 20 to 40 mg for fluvastatin, 20 to 40 to week 54 for atorvastatin compared with an almost mg for lovastatin, and 10 to 20 mg for simvastatin. two-fold increase for fluvastatin and lovastatin. Adverse events experienced by at least 2% of patients Differences in adverse events may be attributed to the were similar at these doses with 12% of atorvastatin, addition of colestipol, since the primary increases in 16% of fluvastatin, 13% of lovastatin, and 9% of simvas­ adverse events were related to the digestive system. tatin patients reporting related adverse events (Table 4). While only 2 atorvastatin patients (2%) had been At week 54, when patients may have received reduc­ exposed to combination therapy by the end of the tase inhibitors and colestipol, the percentage of treat­ study, 57 fluvastatin patients (67%), 20 lovastatin ment-related adverse events experienced by at least 2% patients (24%), and 21 simvastatin patients (24%) were of the test population was the lowest for the atorvastatin given colestipol. In normal primary care situations, an group (14%), which was about half of the percentages increased incidence of related adverse events due to

354 The Journal of Family Practice, Vol. 47, No. 5 (Nov), 1998 TREATING TO MEET NCEP-RECOMMENDED LDL CHOLESTEROL CONCENTRATIONS

TABLE 4 cular events.1'2 In the West of Related Adverse Events, by Body Systems, Experienced by at Least 2% of Patients Scotland Coronary Prevention Atorvastatin Fluvastatin Lovastatin Simvastatin Study, a maximum dose of pravas­ Adverse Event* (n = 86) (n = 85) (n = 86) (n = 87) tatin (40 mg per day) reduced LDL cholesterol by 26% and significantly Week 2 4 t reduced the incidence of myocar­ dial infarction and death from car­ Body as a whole, no. (%) diovascular causes in a primary pre­ Abdominal pain 2(2) 0(0) 2(2) 0(0) vention population.2 Several ongo­ Headache 2(2) 0(0) 0(0) 1 (D ing atorvastatin studies are designed to examine its effects on Digestive system, no, (%) cardiovascular events. In our study, Dyspepsia 2(2) 4(5) 3(3) 0(0) patients treated with atorvastatin Constipation 1 (D 2(2) 0(0) 1 (D achieved a 36% reduction in LDL Flatulence 0(0) 2(2) 1 (D 1 (D cholesterol concentration at the starting dose of the drug, allowing Nervous system more than 70% of patients to meet Insomnia, no. (%) 0(0) 3(3) 1 (D 1 (D their NCEP-recommended LDL cho­ Any event, no. (%) 10(12) 14 (16) 11 (13) 8(9) lesterol goal at this dose. Overall, target LDL cholesterol concentra­ Week 54$ tion was met with less combination therapy and consequently fewer Body as a whole, no. (%) side effects than patients treated Abdominal pain 2(2) 3(4) 2(2) 0(0) with other reductase inhibitors. The Headache 2(2) 0(0) 0(0) 1 (D results of our study indicate that atorvastatin is a highly effective Digestive system, no. (%) treatment for a population at risk Flatulence 1 (D 4(5) 3(3) 2(2) for CHD. Nausea 1 (D 1 (1) 0(0) 2(2) Constipation 2(2) 8(9) 4(5) 3(3) ACKNOWLEDGMENTS Dyspepsia 2(2) 6(7) 6(7) 0(0) This study was funded by Parke-Davis Diarrhea 0(0) 2(2) 1 (D 3(3) Pharmaceutical Research, a division of Rectal hemorrhage 0(0) 2(2) 0(0) 0(0) Warner Lambert Company. The authors wish to thank and acknowl­ edge the other participating investigators: Nervous system, no. (%) Charles Boyajian, MD, Elk Grove Village, Insomnia 0(0) 1 (1) 1 (D 3(3) 111; Alan Brown, MD, Edward Cardio­ Somnolence 0(0) 0(0) 2(2) 0(0) vascular Clinic, Naperville, 111; Stephen Clark, MD, Jacksonville, Fla; Michael Davidson, MD, Chicago Center for Clinical Any event, no. (%) 12 (14) 29 (34) 21 (24) 16(18) Research, Chicago, 111; Richard Guthrie, MD, St. Paul Heart Clinic, St. Paul, Minn; "Considered possibly, probably, or definitely related to treatment or insufficient information. Keith Holten, MD, Providence/UC Family (Patients receiving monotherapy. Center, Cincinnati, Ohio; Kevin Holthaus, (Patients may have added colestipol adjuvant therapy MD, Point Vedra Beach, Fla; William Insull, MD, Lipid Research Clinic, Houston, Tex; Michael Koren, MD, Jacksonville Cardio­ vascular Clinic, Jacksonville, Fla; Charles resins has resulted in high discontinuance rates,14 Margolis, MD, University of Wyoming Family Practice Center, largely because of the occurrence of adverse events. Cincinnati, Ohio; Thomas Novak, MD, West Branch Family Similarly, large discontinuance rates have been report­ Practice Clinic, West Branch, Iowa; Krishna Sikaria, MD, The Heart Center, St. Augustine, Fla; Dance Trence, MD, Group Health ed for these agents from pharmacy prescription refill of Fairview Riverside, Minneapolis, Minn; William Woo, MD, and data obtained in the primary care setting.16 A greater Stephen Moore, MD, Ballard Clinic, Seattle, Wash; and Lawrence percentage of atorvastatin patients met NCEP-recom- Yellen, MD, San Diego, Calif. The authors thank Rachel E. Laskey for assistance with the manuscript, Perry Tresh for study man­ mended target LDL cholesterol concentration with agement, and Dean Smith for his analytical expertise. monotherapy resulting in fewer drug-related adverse events, a lower discontinuance rate, and a slightly REFERENCES higher compliance rate than with those treated with 1. Sacks FM, Pfeffer MA, Moye LA, et al. The effect of pravas­ either fluvastatin, lovastatin, or simvastatin. tatin on coronary events after myocardial infarction in Other studies have shown that reducing LDL cho­ patients with average cholesterol levels. N Eng J Med 1996; 335:1001-9. lesterol significantly reduces incidence of cardiovas­ 2. Shepherd J, Cobbe SM, Ford I, et al. Prevention of coronary

The Journal of Family Practice, Vol. 47, No. 5 (Nov), 1998 3 5 5 TREATING TO MEET NCEP-RECOMMENDED LDL CHOLESTEROL CONCENTRATIONS

heart disease with pravastatin in men with hypercholes­ 1972; 22:98-101. terolemia. N Eng J Med 1995; 333:1301-7. 9. Feskanich D, Buzzard IM, Welch BT, Asp EH, et al. 3. Expert panel on detection, evaluation, and treatment of Comparison of a computerized and manual method of food high blood cholesterol in adults. Summary of the second coding for nutrient intake studies. J Am Diet Assoc 1988' report of the National Cholesterol Education Program 88:1263-7. (NCEP) expert panel on detection, evaluation, and treat­ 10. Steiner PM, Freidel J, Bremner WF, Stein EA. ment of high blood cholesterol in adults (adult treatment Standardization of micromethods for plasma cholesterol, panel II). JAMA 1993; 269:3015-24. triglyceride and HDL-cholesterol with the Lipid Clinics 4. Nawrocki JW, Weiss SR, Davidson MH, et al. Reduction of methodology. J Clin Chem Clin Biochem 1981; 19:850. LDL cholesterol by 25% to 60% in patients with primary 11. Warnick GR, Albers JJ. A comprehensive evaluation of hypercholesterolemia by atorvastatin, a new HMG-CoA the heparin manganese precipitation procedure for esti­ reductase inhibitor. Arterioscler Thromb Vase Biol 1995; mating high density lipoprotein cholesterol. J Lipid Res 15:678-82. 1978; 19:67-76. 5. Davidson MM, McKenney JM, Stein EA, et al. Long-term 12. Lipid and lipoprotein analysis. In: Manual of laboratory efficacy and safety of atorvastatin compared to lovastatin operations: LRC program report (NIH) 75-628. US in hypercholesterolemic patients. Am J Cardiol 1997; Department of Health, Education and Welfare, 1974, Vol 1. 79:1475-81. 13. Koren M, Smith D, Hunninghake D, et al. Cost effective­ 6. Bertolini S, Bittolo Bon G, Campbell LM, et al. The efficacy ness of treatment to National Cholesterol Education and safety of atorvastatin compared to pravastatin in Panel (NCEP) targets with HMG-CoA reductase patients with hypercholesterolemia. Atherosclerosis 1997; inhibitors. PharmacoEconomics. In press. 130:191-7. 14. Schectman G, Hiatt J. Drug therapy for hypercholes­ 7. Remmell PS, Gorder DD, Hall Y, Tillotson JL. Assessing terolemia in patients with cardiovascular disease: factors dietary adherence in the Multiple Risk Factor Intervention limiting achievement of lipid goals. Am J Med 1996' Trial (MRFIT), I: use of a dietary monitoring tool. JAMA 100:197-204. 1980; 76:351-6. 15. Andrade SE, Walker AM, Gottlieb LK, et al. 8. Friedewald WT, Levy RI, Fredrickson DS. Estimation of Discontinuation of antihyperlipidemic drugs—do rates the concentration of low density lipoprotein cholesterol reported in clinical trials reflect rates in primary care set­ in plasma without the use of a centrifuge. Clin Chern tings? N Eng J Med 1995; 332:1125-31.

3 5 6 The Journal of Family Practice, Vol. 47, No. 5 (Nov), 1998